C++ primer 5. ed. A lambda closure in place of a function pointer not allowed?

Question

In C++ primer chapter 14. Function objects:

// ordinary function
int add(int i, int j) { return i + j; }
// lambda, which generates an unnamed function-object class

auto mod = [](int i, int j) { return i % j; };
// function-object class
struct div {
    int operator()(int denominator, int divisor){
        return denominator / divisor;
    }
};

// maps an operator to a pointer to a function taking two ints and returning an int
map<string, int(*)(int,int)> binops;

// ok: add is a pointer to function of the appropriate type
binops.insert({"+", add}); // {"+", add} is a pair § 11.2.3 (p. 426)

However, we can’t store mod or div in binops:

binops.insert({"%", mod}); // error: mod is not a pointer to function

The problem is that mod is a lambda, and each lambda has its own class type. That type does not match the type of the values stored in binops.

But if I use this:

    bin.insert( {"+", [](int x, int y){return x + y;} }); // works fine for me on GCC

Does it belong to C++ standard prior to c++ 20 that prevents using a lambda-expr in place of a pointer to function?

[`binops.insert(mod)` works for me](https://godbolt.org/z/DSEg7_) . Yes, seems to compile with `gcc -std=c++11`. Please post a fulll [MCVE] including all relevant `#include` and `main`. — KamilCuk, Apr 27 '20 at 21:23
also for me, but i'm using C++17, maybe you are using another verison — Alberto Sinigaglia, Apr 27 '20 at 21:24
@Someprogrammerdude: `gcc version 9.3.1 20200317 (Red Hat 9.3.1-1) (GCC)` . — Maestro, Apr 27 '20 at 21:25
That's odd, with GCC 9.3 [it works for me](https://godbolt.org/z/4Egpqt). — Some programmer dude, Apr 27 '20 at 21:27
@Someprogrammerdude: Yea I think so. So you don't think it should work? — Maestro, Apr 27 '20 at 21:27
It's perfectly fin and should work (as it does in my compiler explorer link). — Some programmer dude, Apr 27 '20 at 21:28
You can also use std::function instead of a function pointer, that was actually recommended way to do it. — Strong will, Apr 27 '20 at 21:30
Oh it should work for the lambda, but not for the `div` structure. That's because instances of the `div` structure aren't implicitly convertible to function pointers. You could add such a conversion operator (as is done for the lambdas) though. If you use `std::function` (as mentioned by @Strongwill) it will work fine for the structure as well. — Some programmer dude, Apr 27 '20 at 21:31
@Someprogrammerdude: I've tried it on multiple compilers with multiple C++ version from C++11 up to C++2a and with and without multiple levels of optimizations and it always works! So do you think this is just a mistake in the book? — Maestro, Apr 27 '20 at 21:32
@Someprogrammerdude: Yes correct! it doesn't work for the function-object class `div`. — Maestro, Apr 27 '20 at 21:32
Are you saying that it's the *book* that claims it will not work? Which C++ Primer book are you reading? The one by Stanley Lippman et al, (a highly recommended book), or C++ Primer Plus by Stephen Prata (which is less recommended)? — Some programmer dude, Apr 27 '20 at 21:35
@Someprogrammerdude: C++ primer 5 th. ed by Stanley lippmann. It is said in the book it shouldn't work for lambda. — Maestro, Apr 27 '20 at 21:36
@Maestro The conversion is valid even according to the C++11 Standard [expr.prim.lambda/6], so the book seems to be wrong here. — Daniel Langr, Apr 27 '20 at 21:38
@Maestro BTW, in the 6th edition, this example is no longer there. — Daniel Langr, Apr 27 '20 at 21:42
@Someprogrammerdude: Yes I agree with you. Even there are some errors but still the best book I'm reading. — Maestro, Apr 27 '20 at 21:44
Not the only one, see [errata](http://ptgmedia.pearsoncmg.com/images/9780321714114/errata/9780321714114_errata_10-31-12.html). — Daniel Langr, Apr 27 '20 at 21:45

C++ primer 5. ed. A lambda closure in place of a function pointer not allowed?

0 Answers0